An association between hypoxemia and poor outcomes from traumatic brain injury (TBI) is well documented. However, it is unclear whether hyperoxygenation is beneficial. This registry-based analysis explores the relationship between early hypoxemia and hyperoxemia on outcome from moderate-to-severe TBI. TBI patients (Abbreviated Injury Scale score 3+) were identified from the San Diego County trauma registry. Patients were stratified by arrival partial oxygen pressure (Po(2)) value. Trauma and injury severity score (TRISS) was then used to calculate predicted survival for each patient, with the mean observed-predicted survival differential determined for each arrival Po(2) stratification. Logistic regression was used to quantify the relationship between hypoxemia, hyperoxemia, and outcome from TBI after adjusting for multiple variables including intubation and ventilation status. A total of 3420 patients were included in the analysis. TRISS calculations revealed worse outcomes than predicted for both hypoxemia and extreme hyperoxemia. Logistic regression revealed an optimal Po(2) range (110-487 mm Hg), with an independent association observed between decreased survival and both hypoxemia (OR 0.54; 95% CI 0.42, 0.69; p < 0.001) and extreme hyperoxemia (OR 0.50; 95% CI 0.36, 0.71; p < 0.001). The association between hypoxemia and extreme hyperoxemia and worse outcomes was also present with use of "good outcomes" as the outcome variable (discharge to home, rehabilitation, jail, or psychiatric facility, or leaving against medical advice). We conclude that both hypoxemia and extreme hyperoxemia are associated with increased mortality and a decrease in good outcomes among TBI patients.